| 00:45.18 | *** join/#brlcad stevegt_ (~stevegt@cislunar.TerraLuna.Org) | |
| 04:21.11 | starseeker | grabs opensolaris from virtualboxes.org and tries to get it working... |
| 05:06.36 | starseeker | gah - gcc 3.4.3 |
| 05:16.03 | starseeker | gives sun studio express a shot... |
| 06:10.42 | *** join/#brlcad 52AACLEFP (~stevegt@c-69-181-134-76.hsd1.ca.comcast.net) | |
| 07:53.35 | *** join/#brlcad Ralith (~ralith@S010600221561996a.vc.shawcable.net) | |
| 12:21.24 | brlcad | was a perfectly fine compiler for many years ... should work ;) |
| 12:22.20 | brlcad | so I was right about the savings. reduced a conversion that took about 8 days to around 3 days |
| 12:23.09 | brlcad | not too helpful, but it should also reduce several of the 8 minute -> 3 minute or 80 sec -> 30 sec conversions too |
| 14:32.21 | louipc | sounds like an amazing improvement to me |
| 15:39.07 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 17:31.40 | starseeker | brlcad: which, gcc or sun studio? |
| 17:42.22 | starseeker | old gcc didn't like opennurbs |
| 17:56.39 | ``Erik | how old is old? like 3.x? |
| 17:56.46 | ``Erik | http://www.capstoneturbine.com/news/story.asp?id=536 |
| 17:56.46 | starseeker | 3.4.3 |
| 17:57.31 | ``Erik | (high performance car, 4 big honkin' dc motors, some batteries, and a turbine engine to charge the batteries, 500mi range... awesome) |
| 17:58.05 | starseeker | sweet |
| 17:59.05 | ``Erik | I vagually recall a lot of cruddy c++ code breaking moving to 4.0, the c++ frontend was rewritten and far more strict, wonder if opennurbs depends on things 'being right' that 3 was sloppy on |
| 17:59.52 | ``Erik | (like mebbe using the same symbol in 2 namespaces, 3 was sloppy with namespaces, most people ignored 'em) |
| 18:00.27 | ``Erik | (and the turbine engine used in that badboy is only 40hp, but it can sprint with over 700) |
| 18:04.06 | ``Erik | gcc 3.x can always bootstrap 4.x if that's a need, should probably document that 4.0+ is required somewhere if that holds |
| 18:54.42 | *** join/#brlcad ibot (~ibot@rikers.org) | |
| 18:54.43 | *** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || #brlcad logs: http://ibot.rikers.org/%23brlcad/ || BRL-CAD release 7.16.10 is posted! (20100805) | |
| 19:05.20 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 19:07.30 | *** join/#brlcad velociostrich (~nsd@c-68-37-119-2.hsd1.nj.comcast.net) | |
| 19:40.57 | brlcad | starseeker: "didn't like opennurbs" isn't very descriptive :) |
| 19:41.20 | brlcad | neither is likely going to work out of the box without some changes |
| 19:41.42 | brlcad | we've built with 3.4.3 before, though, so it shouldn't be too hard to fix |
| 19:41.57 | brlcad | sun studio is usually even more picky |
| 19:42.38 | starseeker | opennurbs_brep_region.cpp:1147 error: expected primary-expression before numeric constant |
| 19:43.40 | starseeker | that's the first error from gcc 3.4.3 - I haven't gotten past tcl yet with sun studio - need the sunmath lib and I don't have the test for that wired in yet |
| 19:44.15 | brlcad | that error might imply some sort of #define contstant on subfi |
| 19:44.21 | starseeker | is compiling successfully on OpenSolaris a decent predictor for success with Solaris itself? |
| 19:44.34 | brlcad | yep |
| 19:44.50 | brlcad | not 100% match, but pretty darn close |
| 19:45.01 | starseeker | oh, good - so this isn't a waste :-) |
| 19:46.23 | starseeker | come to think of it, I think clang barfs on opennurbs too |
| 19:46.29 | starseeker | wonder if they're related |
| 19:47.20 | starseeker | decides he might as well do the autotools distcheck as long as he's got this sucker up... |
| 20:14.00 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) | |
| 22:35.39 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) | |
| 23:04.10 | starseeker | blinks - clutils fails, but with an error that looks like it's trying to compile istream.cc from studio express's files? |
| 23:06.05 | starseeker | aaand apparently the Solaris version of yacc doesn't support a -o option |
| 23:07.41 | starseeker | my that's primitive - and a royal pain |
| 23:11.09 | starseeker | recognizes he must handle that case but puts flex and bison in for now... |
| 23:37.22 | starseeker | alright, now that's just flat-out weird - cmake step build succeeds with flex/bison in place |
| 23:39.00 | starseeker | winces - I'm guessing there may be a significant set of patches if we want to get opennurbs compiling with the sun compiler - doesn't even make it past the first file... |
| 23:39.43 | starseeker | "opennurbs_system.h", line 282: Error: A declaration does not specify a tag or an identifier. |
| 23:40.10 | starseeker | "opennurbs_system.h", line 363: Error: #error Verify that long long is a 64 bit integer with your compiler!. |
| 23:48.14 | starseeker | must confess he has no clue what the issue with line 282 is... |
| 23:53.21 | starseeker | but if he adds SUN to the compiler list there and assumes long long for line 363 he gets by those two... to promptly trigger a rash of failures in other headers |
| 23:57.17 | *** join/#brlcad Nohla (~Nohla@201.255.246.105) | |